This program of research is intended to analyze the nature of the coherences that characterize the social behavior of the individual and their links to the perceived dispositions and impressions of consistency formed by the observer. Theoretically, dispositions are reconceptualized in terms of probabifistic relations between types of situations and types of behaviors, rather than as overall behavioral tendencies averaged over multiple situations. Methodologically, the program utilizes a series of field studies, combined with controlled experiments, to test hypotheses about the structure and locus of behavioral consistencies and their perception. The emerging conditional reconceptualization of behavioral dispositions will be applied to a reasonably diverse sample of domains and populations to allow an evaluation of its generality and limits. Aggression, withdrawal, and prosocial behavior are the domains studied in a population with social adjustment problems, aged 6 to 17 years. Parallel studies also are conducted on conscientiousness and friendliness with college and private school students. Data from our already in-progress behavioral archives on consistency, and from new studies, provide multiple behavior samples for each subject across situations, as well as dispositional judgments, and perceptions about the situations sampled and the behaviors they require. Relevant moderator variables are assessed, with a focus on the role of competencies, both as demanded by the situation and as available to the person. Individuals judged reliably as good (prototypic) exemplars of different dispositions are compared to identify the diagnostic if then contingencies distinctive for them. Comparisons will include characteristic diagnostic contingencies obtained from intuitive descriptions of these subjects, and from observations of if then regularities in their ongoing behavior as it unfolds. Drawing on the field studies, experiments will test the predictive significance and operation of the diagnostic contingencies identified for particular dispositions. In these experiments, perceivers are exposed to information about targets actual behavior but the diagnostic contingencies are manipulated experimentally to test the effects on 1. The nature of the impressions generated in the perceiver, and 2. their utility for the accurate prediction of the targets' actual behavior as observed and judged in the field studies. Laboratory studies also will test differences between the dispositional exemplars in the accessibility of dispositionally-relevant behavior scripts, expectations, and beliefs when they are primed by exposure to the appropriate situations. A follow-up wave in our longitudinal study of self regulatory competencies will test hypotheses about specific diagnostic contingencies assessed two decades ago in preschool as predictors of the subjects' current competencies and dispositional status.